Selection of optimum gear ratios - a discussion

[Shaun_E]

Following this thread I started to think about how optimum gearing is decided. Over the years I have read a number of posts that discuss the merits of 5 speed vs 6 speed, close ratio boxes, re-ratioed 5 speeds, changing gearing for different race circuits, but I have never completely understood. My basic understanding is that when you reach revs for peak power (or maybe just over) then you change gear and you want the drop in revs to be such that you fall somewhere near peak torque - is this correct? This would suggest that power band is an important parameter. What kind of parameters would influence the gearing choice - not just the difference in ratios but also the maximum and minimum?

It seems to me that the more power a car develops the higher the overall gearing (to achieve max speed?) yet rally cars appear to have low gearing - does this improve acceleration? Also drag race cars tend to run lower final drives than their street equivalents - for the same reason?

Please let me know your thoughts and any links to good explanations.

Thanks,

Shaun

 

 

Yellow SL #32

[Sheds Moderator]

Yes, this is true. Lower gearing allows the engine to be spinning faster at any given road speed so developing more power (up to a point). This is why your car accelerates faster in lower gears than higher gears. Close ratio boxes allow you to have the car in whatever gear is best for max torque and powwer at all speeds.

 

I have an article in French re gear ratios in 2CVs and the theory behind it. I'm happy to copy and post it if you can plough through the text and diagrams.

[Graham Sewell]

With the exception of motorway cruising, my understanding (which is very simplified) is that at max engine power you can predict a max speed based on aerodynamics and rolling resistance. This (with a 1:1 top gear) gives the ratio for the diff.

 

Then at the other extreme, you look at what gearing is needed to pull away smoothly (if briskly), and then sprinkle the remaining 2/3/4 gears between these two extremes.

 

However, when circuit racing you can start to optimise gears for specific corners so that you are on the power band at the apex of the corner. I remember our instructor at Prescott School saying that his gearbox car has two second gears for different corners and no "third" - so depending on the corner he would select one of his two second gears before going for fourth.

 

Low tech luddite - xflow and proud!

[Normans_Ghost]

Somewhat similar to Harolds question about power. The engine charateristics and useage determine number of gears required. An engine with a narrow power band will be quicker* with lots of close ratio's because the engine speed can be kept within this narrow power band. If you have a wide power band then you need less ratio's to keep the engine speed "on the boil". Some would say more is best but why waste valueable seconds changing gear when you don't have to. Modern saloon cars are fitted with 6 speed boxes for a different reason - to get better MPG figures. In principle lower ratios will give you quicker acceleration over top speed. We could all gear our cars to get to higher top speed but at the cost of taking longer to get there.

 

Hope that helps, but I'm sure someone will disagree.

 

*quicker=acceleration, faster=top speed

 

Norman Verona, 1989 BDR 220bhp, Mem No 2166, the full story here

[neil.cavanagh]

if you were picking the ratios for a specific track wouldn't the decision be something like this..

 

what is the fastest speed I can get upto on this particular track?

 

Then set top gear so you would be just about bouncing off the rev limiter at this speed.

 

then spread the gears between with 1st being short enough that you can pull off OK still.

 

After that you would then look at fiddling with individual ratios to suit bends/ straights (as in if I lengthen 2nd slightly I won't have to change up to 3rd and straight back to 2nd when going from bend a to bend B).

 

You'll probably then find your going a lot quicker and need to raise top gear as your spending too long on the limit... so you then need to re-work the mid range gears to be optimum...

 

For the road the decision is differnt... how many gears can I have.

then compromise between top gear being nice and long for motorway cruising, and 1st gear being useable in traffic, and then spread the other gears around, maybe making sure 2nd gear can get you to 62mph for the 0-60 stat..

 

[Shaun_E]

To keep things simple (ish) how would you choose ratios for the quickest straight line acceleration (e.g. drag racing or the traffic lights grand prix)? So quick 0-60 or quick quarter mile times. I know that traction is an issue from launch and perhaps that affects the decision on first gear?

Keep 'em coming. Wheres PC when you need him?

 

Yellow SL #32

[Peter Carmichael]

Going back to the original post...

 

when you reach revs for peak power (or maybe just over) then you change gear

Definitely "after" peak power if the engine is happy to turn the revs.

 

and you want the drop in revs to be such that you fall somewhere near peak torque - is this correct?

 

No. This isn't correct. Basically the higher the revs after the gear change the faster you will accelerate.

 

For drag racing (a race over a fixed distance from a standing start), it is most important to gain as much speed as you can early on in the process. The launch is everything. The first gearchange should be as late as possible, but between launch and the first gearchange you should be on the verge of breaking traction.

 

For circuit racing, you basically want a ratio to accelerate you out of each corner and you want to be *just coming onto the limiter* at the end of each significant straight.

[Shaun_E]

Basically the higher the revs after the gear change the faster you will accelerate.

So is it only gear change time that stops us having more gears? I guess a CVT would allow maximum revs (or peak power revs) at any speed and therefore would provide the ultimate acceleration.

Why do people say that a 5 speed box is better with a "torquey" engine? Is this bullsh*t or does the increased torque or wider power band and reduced number of gear changes overcome the benefits of closer ratio gears?

Are most Caterhams geared for maximum speed at peak revs in top gear? Would a 3.9 or 4.1 diff ratio make the car more accelerative (at the expense of top speed) or do other factors come into play? With the introduction of DSG gearboxes, would more closer ratio gears be a good thing?

 

 

Yellow SL #32

 

Edited by - Shaun_E on 14 Oct 2004 15:47:08

[PaulD]

Shaun,

 

Finding the optimum gear ratios for straight line acceleration would make a nice introduction to vehicle dynamics simulation.

 

Start with this concept: You have a data logger in the car and during the acceleration run, you measure RPM vs time. If you have a power curve, you can convert the RPM/time curve to a power / time curve. To get the best performance from the car, you need to maximise the area under the power/time curve.

 

To make the simulation, you need 3 pieces of data; the mass of the car, a power curve (at the flywheel) and a coast down run.

 

The coast down run will give you the losses due to aero and rolling resistance. To do it, find a level runway (or quiet, flat motorway), take the car up to speed, put it out of gear and record the time as the speed to drops off. Since you know the mass of the car, From your data, you can use F=ma to get a plot of retarding force/speed.

 

With this data, you can program a simulation in VB or whatever. For any road speed point, you will know the engine speed (from the gear ratio), you will know the power (from the power curve) and you will know the resistance (from your coast -down run). using P = ma - R, you can predict the acceleration and, hence, the speed at the next time slice.

 

I suspect that the simulation will show that you need the higher gears to be closer together than lower gears.

 

Happy programing

 

Paul

[Shaun_E]

Paul - sounds like a project for another day - i.e. once my car has been completely dismantled and rebuilt!. Thanks for the theory though - it all adds to my understanding. Really I'm just trying to get a scientific angle on all the banter about torquey engines vs high revving engines and 5 speed for touring vs 6 speed for track and all the other *cough*bull that gets bandied around. I like to understand what I'm talking about from 1st principles rather than just gut feeling.

I really like a revvy engine and my last couple of tintops have had 6-speed boxes as does my Caterham. I don't really understand why anyone would want a "lazy" engine in a 7 - to me half of the fun is keeping it on the boil. If I wanted loads of torque I'd get a big V8 (Corvette maybe 😳).

Cheers,

Shaun

 

Yellow SL #32